Project : evasion
Section: New Results
Morphogenesis, expansion textures
The purpose is to offer a modelling tool for complex surfaces which shape results from a growing phenomena (e.g. biological or geological surfaces), by allowing the user to control the growing rather than the shape itself. Controlling is done through a texture encoding the intensity and orientation of the deformations, either explicitly (e.g. map of rifts and subductions) or generatively (e.g. reticulation, hot spots...). Moreover, the control can also be interactive by `painting' the effects directly on the surface. In particular, we simulated wrinkles and folds, which adds a complexity that is important for realism, but that the artist wants to control without having to explicitly describe the local shape (voir figure 8).
Jean Combaz is finishing his PhD on Expansion Textures. He is expected to defend his PhD in early 2004.
Efficiently animating virtual clay is a challenge, since neither optimisations proposed for solids (and based on a constant topology) nor for fluids (since there is a moving limit surface) are directly applicable. We, thus, proposed the first real-time model for this material , based on a layered approach. Three sub-models respectively handling large-scale deformations, local matter displacements, and surface tension, cooperate overtime for providing the desired behaviour.
Physically-based interactive rigid bodies
We addressed the issue of simulating large cliff collapses, in collaboration with the laboratory 3S (Sols, Solids, Structures) whose specialisation is mechanics. We have currently worked on two main topics:
extend a popular mechanical model based on individual spheres, namely the Discrete Element Model, to assemblies of spheres in order to obtain a better rotational behaviour;
apply levels of detail to collision detection and modelling.
Highly colliding deformable bodies
We addressed the issue of simulating highly deformable objects in real-time, such as human tissues or cloth. The main problem is to detect and handle multiple (self-)collisions within the bodies. We developed a new approach for collision detection, based on a pool of "active pairs" of geometric primitives  . These pairs are randomly chosen, and they iteratively converge to a local distance minimum or to a pair of colliding elements. Managing the size of the pool allows us to tune the computation time devoted to collision detection. Temporal coherence is obtained by reusing the interesting pairs from one step to another. While not guaranteeing that we detect every collision at each time step, this approach has shown its efficiency when applied to the simulation of an intestinal surgery simulator as one can see in figure 9 (see also sections 6.6.3 and 8.2.2 for the resulting medical application).
Parallel simulation of cloth
We addressed the issue of simulating cloth on a PC cluster, in collaboration with the laboratory ID and the company Yxendis. Cloth is modelled as a physically-based deformable mesh. There are two important difficulties:
to reduce the amount of communication between the cluster nodes;
to transfer the data from the cluster to the rendering machine.
We split the cloth in compact pieces, thus reducing communication to data related to the borders of the patches. We use socket communications to transfer the data scattered in the cluster to the rendering machine   .